WSJ (Einstein 3) – Wash Post (Secret Sauce) – NSA West? (Camp Williams, UT) – OMB M-07-16 (Breach of PI Info) – FISMA Reporting Requirements


Troubles Plague Cyberspy Defense

The flagship system designed to protect the U.S. government’s computer networks from cyberspies is being stymied by technical limitations and privacy concerns, according to current and former national-security officials.

NSA-HQbldgThe latest complete version of the system, known as Einstein, won’t be fully installed for 18 months, according to current and former officials, seven years after it was first rolled out. This system doesn’t protect networks from attack. It only raises the alarm after one has happened.

A more capable version has sparked privacy alarms, which could delay its rollout…

“The good news is, I think [the administration] appears to be taking a close look at how best to do this,” Mr. Kurtz said. “The bad news is, while they work to figure it out, the security of our networks is not necessarily getting any better.”

Homeland Security spokeswoman Amy Kudwa described the various rollouts as “incremental improvements” designed also to protect privacy and civil liberties. “We don’t want to let the perfect be the enemy of the good,” she said…

It will take 18 months to launch Einstein 2 across most of the government, a senior Homeland Security official said, and then 96 smaller agencies will follow. Plans are already under way for Einstein 3. As envisioned by the Bush administration, Einstein 3 would draw from an NSA program that automatically identifies and deflects security breaches, according to former officials familiar with the program….

The three phases of the cyber-detection program dubbed Einstein.

  • Einstein 1: Monitors Internet traffic flowing in and out of federal civilian networks. Detects abnormalities that might be cyber attacks. Is unable to block attacks.
  • Einstein 2: In addition to looking for abnormalities, detects viruses and other indicators of attacks based on signatures of known incidents, and alerts analysts immediately. Also can’t block attacks.
  • Einstein 3: Under development. Based on technology developed for a National Security Agency program called Tutelage, it detects and deflects security breaches. Its filtering technology can read the content of email and other communications.


Cybersecurity Plan to Involve NSA, Telecoms

The Obama administration will proceed with a Bush-era plan to use National Security Agency assistance in screening government computer traffic on private-sector networks, with AT&T as the likely test site, according to three current and former government officials.

— “We absolutely intend to use the technical resources, the substantial ones, that NSA has. But . . . they will be guided, led and in a sense directed by the people we have at the Department of Homeland Security,” the department’s secretary, Janet Napolitano, told reporters in a discussion about cybersecurity efforts.

… The program is the most controversial element of the $17 billion cybersecurity initiative the Bush administration started in January 2008. Einstein 3 is crucial, advocates say, in an era in which hackers have compromised computer systems at the Commerce and State departments and have taken military jet data from a defense contractor….

“That’s the secret sauce,” one official said. “It’s the stuff they have that the private sector doesn’t.”  But it is also the prospect of NSA involvement in cybersecurity that fuels concerns about unwarranted government snooping into private communication…

…The pilot program has two goals. The first is to prove that the telecommunications firm can route only traffic destined for federal civilian agencies through the monitoring system. The second is to test whether the technology can work effectively on civilian government networks. The sensor box would scan e-mail messages and other content just before they enter the civilian agency networks.

The classified NSA system, known as Tutelage, has the ability to decide how to handle malicious intrusions — to block them or watch them closely to better assess the threat, sources said. It is currently used to defend military networks….

Deseret News By Lee Davidson and Amy Joi O’Donoghue

Published: Friday, July 3, 2009

Utah will host new $1.9 billion NSA spy center

So the agency began looking for ways to decentralize operations. In 2007, it announced plans to build a second data center in San Antonio. The Utah center will now be the NSA’s third.

The Camp Williams site is near major electrical power-transmission lines that serve the Wasatch Front and sits on 28,000 acres straddling the border of Utah and Salt Lake counties.

The regional center is the training base to the only military unit of its kind in the world, a linguistics unit 1,200-strong called the 300th Military Intelligence Brigade. About 600 members are from Utah, which McIntire says bolsters its ranks because of the unique population of returned LDS missionaries who can speak another language.Camp-Williams

Camp Williams already enjoys a steadfast relationship with other branches of the military, local police and the community in general. Many church groups make use of the barracks and obstacle course for events, and police agencies hold emergency-vehicle training there.

However, the Baltimore Sun reported in 2006 that the operations at Fort Meade had maxed out the electric capacity of the Baltimore area’s power grid — and the NSA was then unable to install some new supercomputers for fear of blowing out the electrical infrastructure of the area.

Spies like us: NSA to build huge facility in Utah

The Salt Lake TributeBy Matthew D. LaPlante 07/02/2009

Officials familiar with the project say it may bring as many as 1,200 high-tech jobs to Camp Williams, which borders Salt Lake, Utah and Tooele counties.

It will also require at least 65 megawatts of power — about the same amount used by every home in Salt Lake City combined. A separate power substation will have to be built at Camp Williams to sustain that demand, said Col. Scott Olson, the Utah National Guard’s legislative liaison.

He noted that there were two significant power corridors that ran though Camp Williams — a chief factor in the NSA’s desire to build there.


OMB M-07-16, Safeguarding Against and Responding to the Breach of Personally Identifiable Information (May 22, 2007) (22 pp. PDF) – Memorandum released by the Office of Management and Budget to the heads of executive departments and agencies within the Federal Government requiring Federal agencies to review their data holdings and ensure data quality requirements are being met.

Sensitive Database Extracts Technical Frequently Asked Questions

This Frequently Asked Questions (FAQ) document addresses technical aspects associated with implementing the Office of Management and Budget (OMB) requirement to log and verify sensitive database extracts, which was required by OMB Memorandum M-07-16, “Safeguarding Against and Responding to the Breach of Personally Identifiable Information” which reiterates the log and verify requirement set forward in M-06-16, “Protection of Sensitive Agency Information,” issued in June 2006. Topics covered in this FAQ include data extract logging, restrictions, verification, and erasure.


1. What is the requirement in the OMB memorandum?

The text of the requirement, as stated on page 7 of OMB M-07-16, is “Log all computer-readable data extracts from databases holding sensitive information and verify each extract, including whether sensitive data has been erased within 90 days or its use is still required.”

2. What is a computer-readable data extract from a database?

This involves retrieving data from a database through a query and saving the data into a separate computer-readable entity such as another database, a spreadsheet, or a text file.

3. What types of information does the requirement apply to?

Although much of M-07-16 focuses on personally identifiable information (PII), the log and verify requirement apples to all sensitive information, including sensitive PII.

4. What is the purpose of the requirement?

The purpose of the requirement is to ensure that data extracts containing sensitive information are erased when they are no longer needed. This reduces the likelihood of sensitive information being breached.


5. Which data extracts need to be logged?

All data extracts from databases that contain sensitive information need to be logged.

6. What information should be logged for each extract?

NIST Special Publication (SP) 800-53 Revision 2, Recommended Security Controls for Federal Information Systems, specifies an Audit and Accountability (AU) family of technical security controls, which encompasses audit logging requirements. Control number AU-3, Content of Audit Records, states that “audit record content includes, for most audit records: (i) date and time of the event; (ii) the component of the information system (e.g., software component, hardware component) where the event occurred; (iii) type of event; (iv) user/subject identity; and (v) the outcome (success or failure) of the event.” In addition to logging this information for each extract, agencies may also log other types of information. For example, agencies may log whether each data extract contains sensitive information, for future use in determining which extracts need to be erased. Agencies may also describe the purpose and length of time for which extracted sensitive information will be used.

7. What recommendations does NIST provide for logging?

In addition to the audit logging-related security controls specified in NIST SP 800-53 Revision 2, NIST has developed SP 800-92, Guide to Computer Security Log Management. SP 800-92 provides recommendations for developing, implementing, and maintaining log management practices throughout an enterprise.


8. How can my agency reduce the number of data extracts that are subject to the requirement?

This can be accomplished by reducing the amount of sensitive information, including sensitive PII, in its databases and by limiting users’ ability to perform extracts from databases with sensitive information.

9. What are some examples of how an agency can reduce the amount of sensitive PII in its databases?

As stated in OMB M-07-16, agencies must collect and retain only the minimum sensitive PII necessary. Agencies may also use data scrubbing techniques to remove sensitive information from database records. Data scrubbing can remove sensitive information permanently, such as replacing PII values with pseudonyms that provide the ability to sort and quantify populations as groups but not individuals. Data scrubbing can also remove PII temporarily, such as mapping PII values to pseudonyms, storing the mappings in a separate file, and replacing the PII values in the database with the pseudonyms. Only an individual with access to both the database and the mapping file could match the individuals’ actual identities with the corresponding database records.

10. How can an agency limit users’ ability to perform data extracts from databases with sensitive information?

Agencies may grant only authorized users the least access necessary to such databases and to the sensitive information within each database. This could include restricting the types of queries that users can perform and the database fields (for example, social security number) that users can view and include in extracts. Another method is to permit users to access sensitive information in databases only through applications that tightly restrict the users’ access to the sensitive information, instead of permitting direct database access. Such applications could manage the data extract process by permitting extracts only when necessary, scrubbing sensitive information, such as sensitive PII, during extraction, forcing all extracts containing sensitive information to be stored centrally, and interacting with centrally-stored extracts on behalf of users so that the users cannot directly access extracts. Agencies may also use other options for limiting data extracts.

11. What technical methods are available for restricting where sensitive extracts are stored?

In addition to the application-based method mentioned above, there are other methods that agencies may use to limit where sensitive extracts are stored. For example, agencies may configure their remote access solutions so as not to permit access to sensitive information databases from mobile devices and non-organization computers (e.g., personally-owned home computers). Agencies could also permit extracts to be stored only on media protected by storage encryption technology. Other methods are more complex and may require considerable planning and deployment time. One example is requiring that sensitive extracts be stored within and accessed only through encrypted virtual machines, which may be set to expire after 90 days. Another example is implementing centralized processing for access to sensitive databases, where the data never leaves the centralized servers and the applications that access the data are run only through thin client solutions.


12. What is required for verifying a sensitive extract?

Agencies may accomplish extract verification through simple checks. An example of such a solution is ad hoc attestation. This involves implementing one or more systems to log the creation of extracts containing sensitive information and to send each extractor a message after 90 days that requires that the extractor either attests to having erased the extract or justifies why the extract is still needed. Agencies may implement more rigorous and formal verification processes than ad hoc attestation to achieve greater confidence in extracts being erased. An example of a more rigorous verification process is storing all extracts on a well-secured centralized system, prohibiting users from directly accessing the extracts, and running a utility that automatically erases extracts 90 days after creation.

13. What is required for erasing a sensitive extract?

The actions needed to erase an extract vary based on the system or media where the extract has been stored. For example, erasing an extract stored on read-only removable media may necessitate physical destruction of the removable media, whereas erasing an extract on a centralized server may involve deleting the extract file and logically sanitizing the portions of the server media that held the file, as well as ensuring that all copies of the extract are properly erased from server backups. Data artifacts from extracts, such as temporary files, may also need to be erased. The procedures for erasing sensitive extracts can result in a significant operational impact on agencies.

14. What other types of technical solutions could be used for sensitive extract verification and erasure?

In addition to the solutions described above, agencies can also implement long-term solutions that automate most of the verification and erasure processes, thus reducing operational impact. Such solutions generally require at least a few years’ effort to implement, so agencies that choose to implement one or more of the long-term solutions may implement one or more of the currently available solutions described above in the meantime. Examples of possible long-term solutions are as follows:

Use a trusted Digital Rights Management (DRM) platform or similar solution to manage extracts. Such technologies could be used to permit access to each extract for a certain number of days and by particular users, as well as to restrict how each extract can be used (e.g., preventing an extract from being copied or printed). Designing and implementing scalable DRM-type infrastructures and supporting systems for database extract management, including the deployment of client and server applications and platforms that support the chosen technology, is likely to require significant time and resources (at least two years).

Implement centralized processing for access to sensitive databases using dumb terminals. This is similar to the thin client solution described earlier, except that the dumb terminals have no memory or storage, which prevents any data from being stored locally. Today’s versions of “dumb terminals” are actually emulations that run on general-purpose computers, which means that sensitive data could be stored locally. This solution cannot be implemented on a large scale in the near term using current off-the-shelf components.

Automatically encrypt each extract, centrally manage all the keys, and destroy the keys at the appropriate times to expire the extracts. Identity-based cryptography could extend this scheme to provide finer-grained access control. These methods are currently in the research stage and cannot be implemented in the near term.






May 22, 2007



FROM: Clay Johnson III

Deputy Director for Management

SUBJECT: Safeguarding Against and Responding to the Breach of Personally Identifiable Information

Safeguarding personally identifiable information1 in the possession of the government and preventing its breach are essential to ensure the government retains the trust of the American public. This is a responsibility shared by officials accountable for administering operational and privacy and security programs, legal counsel, Agencies’ Inspectors General and other law enforcement, and public and legislative affairs. It is also a function of applicable laws, such as the Federal Information Security Management Act of 2002 (FISMA)2 and the Privacy Act of 1974.3123 Safeguarding personally identifiable information in the possession of the government and preventing its breach are essential to ensure the government retains the trust of the American public. This is a responsibility shared by officials accountable for administering operational and privacy and security programs, legal counsel, Agencies’ Inspectors General and other law enforcement, and public and legislative affairs. It is also a function of applicable laws, such as the Federal Information Security Management Act of 2002 (FISMA) and the Privacy Act of 1974.

As part of the work of the Identity Theft Task Force,4 this memorandum requires agencies to develop and implement a breach5 notification policy6 within 120 days As part of the work of the Identity Theft Task Force, this memorandum requires agencies to develop and implement a breach notification policy 4567 within 120 days. The attachments to this memorandum outline the framework within which agencies must develop this breach notification policy while ensuring proper safeguards are in place to protect the information. Agencies should note the privacy and security requirements addressed in this Memorandum apply to all Federal information and information systems.8 Breaches subject to notification requirements include both electronic systems as well as paper documents. In short, agencies are required to report on the security of information systems in any formant (e.g., paper, electronic, etc.). 9

In formulating a breach notification policy, agencies must review their existing requirements with respect to Privacy and Security (see Attachment 1). The policy must include existing and new requirements for Incident Reporting and Handling (see Attachment 2) as well as External Breach Notification (see Attachment 3). Finally, this document requires agencies to develop policies concerning the responsibilities of individuals authorized to access personally identifiable information (see Attachment 4).

Within the framework set forth in the attachments, agencies may implement more stringent policies and procedures reflecting the mission of the agency. While this framework identifies a number of steps to greatly reduce the risks related to a data breach of personally identifiable information, it is important to emphasize that a few simple and cost-effective steps may well deliver the greatest benefit, such as:

o reducing the volume of collected and retained information to the minimum necessary;

o limiting access10 to only those individuals who must have such access; and

o using encryption, strong authentication procedures, and other security controls to make information unusable by unauthorized individuals.

This Memorandum should receive the widest possible distribution within your agency and each affected organization and individual should understand their specific responsibilities for implementing the procedures and requirements. Materials created in response to this Memorandum and attachments should be made available to the public through means determined by the agency, e.g., posted on the agency web site, by request, etc.

Consistent with longstanding policy requiring agencies to incorporate the costs for securing their information systems, all costs of implementing this memorandum, including development,

implementation, notification to affected individuals, and any remediation activities, will be addressed through existing agency resources of the agency experiencing the breach.

Because of the many alternate ways to implement a risk-based program within the framework provided, this Memorandum, or its attachments, should not be read to mean an agency’s failure to implement one or more of the many security provisions discussed within11 would constitute less than adequate protections required by the Privacy Act. These new requirements do not create any rights or benefits, substantive or procedural, which are enforceable at law against the government.

Questions about this Memorandum should be directed to Hillary Jaffe of my staff at


Attachment 1: Safeguarding Against the Breach of Personally Identifiable Information

This Attachment reemphasizes the responsibilities under existing law, executive orders, regulations, and policy to appropriately safeguard personally identifiable information and train employees on responsibilities in this area (Section A).12 It also establishes two new privacy requirements and discusses five security requirements as described below (Sections B and C).

A. Current Requirements

1. Privacy Act Requirements. In particular, the Privacy Act of 1974 (Privacy Act)13 requires each agency to:

a. Establish Rules of Conduct. Agencies are required to establish “rules of conduct for persons involved in the design, development, operation, or maintenance of any system of records, or in maintaining any record, and instruct each such person with respect to such rules and the requirements of [the Privacy Act], including any other rules and procedures adopted pursuant to [the Privacy Act] and the penalties for noncompliance.” (5 U.S.C. § 552a(e)(9))

b. Establish Safeguards. Agencies are also required to “establish appropriate administrative, technical, and physical safeguards to insure the security and confidentiality of records and to protect against any anticipated threats or hazards to their security or integrity which could result in substantial harm, embarrassment, inconvenience or unfairness to any individual on whom information is maintained.” 14

c. Maintain accurate, relevant, timely and complete information.

The Privacy Act also requires personally identifiable information within a system of records to be maintained in a manner that is accurate, relevant, timely, and complete including through the use of notices to the public.

It is important for agencies to fulfill their responsibilities with respect to identifying systems of records and developing and publishing notices as required by the Privacy Act and OMB’s implementing policies.

By collecting only the information necessary and managing it properly, agencies can often reduce the volume of information they possess, the risk to the information, and the burden of safeguarding it.

2. Security Requirements.

Below are four particularly important existing security requirements agencies already should be implementing:

a. Assign an impact level to all information and information systems. Agencies must follow the processes outlined in Federal Information Processing Standard (FIPS) 199, Standards for Security Categorization of Federal Information and Information Systems, to categorize all information and information systems according to the standard’s three levels of impact (i.e., low, moderate, or high). Agencies should generally consider categorizing sensitive personally identifiable information (and information systems within which such information resides) as moderate or high impact.

b. Implement minimum security requirements and controls. For each of the impact levels identified above, agencies must implement the minimum security requirements and minimum (baseline) security controls set forth in FIPS 200, Minimum Security Requirements for Federal Information and Information Systems, and NIST Special Publication 800-53, Recommended Security Controls for Federal Information Systems, respectively.

c. Certify and accredit information systems. Agencies must certify and accredit (C&A) all information systems supporting the operations and assets of the agency, including those provided or managed by another agency, contractor, or other source.17 The specific procedures for conducting C&A are set out in NIST Special Publication 800-37, Guide for the Security Certification and Accreditation of Federal Information Systems, and include guidance for continuous monitoring of certain security controls. Agencies’ continuous monitoring should assess a subset of the management, operational, and technical controls used to safeguard such information (e.g., Privacy Impact Assessments).

d. Train employees. Agencies must initially train employees (including managers) on their privacy and security responsibilities before permitting access to agency information and information systems. Thereafter, agencies must provide at least annual refresher training to ensure employees continue to understand their responsibilities.18 Additional or advanced training should also be provided commensurate with increased responsibilities or change in duties.

Both initial and refresher training must include acceptable rules of behavior and the consequences when the rules are not followed. For agencies implementing tele-work and other authorized remote access programs, training must also include the rules of such programs.19

B. Privacy Requirements

1. Review and Reduce the Volume of Personally Identifiable Information.

a. Review Current Holdings. Agencies must now also review their current holdings of all personally identifiable information and ensure, to the maximum extent practicable, such holdings are accurate, relevant, timely, and complete, and reduce them to the minimum necessary for the proper performance of a documented agency function.20 Agency-specific implementation plans and progress updates regarding this review will be incorporated as requirements in agencies’ annual report under FISMA.

Following this initial review, agencies must develop and make public a schedule by which they will periodically update the review of their holdings. This schedule may be part of an agency’s annual review and any consolidated publication of minor changes of Privacy Act systems of records notices.

To help safeguard personally identifiable information, agencies are reminded they must meet the requirements of FISMA and associated policies and guidance from the OMB and NIST.21 FISMA requires each agency to implement a comprehensive security program to protect the agency’s information and information systems; agency Inspectors General must independently evaluate the agency’s program; and agencies must report annually to OMB and Congress on the effectiveness of their program.

Within the above framework, agencies may implement more stringent procedures governed by specific laws, regulations, and agency procedures to protect certain information, for example, taxpayer data, census information, and other information.

2. Reduce the Use of Social Security Numbers.

a. Eliminate Unnecessary Use. Agencies must now also review their use of social security numbers in agency systems and programs to identify instances in which collection or use of the social security number is superfluous. Within 120 days from the date of this memo, agencies must establish a plan in which the agency will eliminate the unnecessary collection and use of social security numbers within eighteen months.22

b. Explore Alternatives. Agencies must participate in government-wide efforts to explore alternatives to agency use of Social Security Numbers as a personal identifier for both Federal employees and in Federal programs (e.g., surveys, data calls, etc.).

C. Security Requirements

While agencies continue to be responsible for implementing all requirements of law and policy, below are five requirements23 agencies must implement which derive from existing security policy and NIST guidance. These requirements are applicable to all Federal information, e.g., law enforcement information, etc.

• Encryption. Encrypt, using only NIST certified cryptographic modules, 24 all data on mobile computers/devices carrying agency data unless the data is determined not to be sensitive, in writing, by your Deputy Secretary25 or a senior-level individual he/she may designate in writing;

• Control Remote Access. Allow remote access only with two-factor authentication where one of the factors is provided by a device separate from the computer gaining access;

• Time-Out Function. Use a “time-out” function for remote access and mobile devices requiring user re-authentication after thirty minutes of inactivity;

• Log and Verify. Log all computer-readable data extracts from databases holding sensitive information and verify each extract, including whether sensitive data has been erased within 90 days or its use is still required; and

• Ensure Understanding of Responsibilities. Ensure all individuals with authorized access to personally identifiable information and their supervisors sign at least annually a document clearly describing their responsibilities.

Agencies should also contemplate and incorporate best practices to prevent data breaches. Examples of such practices might include using privacy screens when working outside the office or requiring employees to include laptop computers in carry-on luggage rather than checked baggage.

Attachment 2: Incident Reporting and Handling Requirements

This Attachment applies to security incidents involving the breach of personally identifiable information whether in electronic or paper format. For the purposes of reporting, agencies must continue to follow existing requirements, as modified and described below.

A. Existing Requirements

1. FISMA Requirements. FISMA requires each agency to:

• implement procedures for detecting, reporting and responding to security incidents, including mitigating risks associated with such incidents before substantial damage is done

• notify and consult with:

o the Federal information security incident center

o law enforcement agencies and Inspectors General

o an office designated by the President for any incident involving a national security system

o any other agency or office in accordance with law or as directed by the President.26

• implement NIST guidance and standards27

Federal Information Processing Standards Publication 200 (FIPS 200) and NIST Special Publication 800-53 provide a framework for categorizing information and information systems, and provide minimum security requirements and minimum (baseline) security controls for incident handling and reporting. The procedures agencies must already use to implement the above FISMA requirements are found in two primary guidance documents: NIST Special Publication 800-61, Computer Security Incident Handling Guide28; and the concept of operations for the Federal security incident handling center located within the Department of Homeland Security, i.e., United States Computer Emergency Readiness Team (US-CERT).29

2. Incident Handling and Response Mechanisms. When faced with a security incident, an agency must be able to respond in a manner protecting both its own information and helping to protect the information of others who might be affected by the incident. To address this need, agencies must establish formal incident response mechanisms. To be fully effective, incident handling and response must also include sharing information concerning common vulnerabilities and threats with those operating other systems and in other agencies. In addition to training employees on how to prevent incidents, all employees must also be instructed in their roles and responsibilities regarding responding to incidents should they occur.

B. Modified Agency Reporting Requirements

1. US-CERT Modification. Agencies must report all incidents involving personally identifiable information to US-CERT. This reporting requirement does not distinguish between potential and confirmed breaches. The US-CERT concept of operations for reporting Category 1 incidents is modified as follows:

Category 1. Unauthorized Access or Any Incident Involving Personally Identifiable Information. In this category agencies must report when: 1) an individual gains logical or physical access without permission to a federal agency network, system, application, data, or other resource; or 2) there is a suspected or confirmed breach of personally identifiable information regardless of the manner in which it might have occurred. Reporting to US-CERT is required within one hour of discovery/detection.

• For incidents involving personally identifiable information, agencies must:

o Continue to follow internal agency procedures for notifying agency officials including your agency privacy official and Inspector General;

o Notify the issuing bank if the breach involves government-authorized credit cards; and

o Notify US-CERT within one hour. Although only limited information about the breach may be available, US-CERT must be advised so it can assist in coordinating communications with the other agencies. Updates should be provided as further information is obtained.

• Under specific procedures established for these purposes, after notification by an agency, US-CERT will notify the appropriate officials.

• Monthly, US-CERT will distribute to designated officials in the agencies and elsewhere, a report identifying the number of confirmed breaches of personally identifiable information and will also make available a public version of the report.

2. Develop and Publish a Routine Use.

a. Effective Response. A federal agency’s ability to respond quickly and effectively in the event of a breach of federal data is critical to its efforts to prevent or minimize any consequent harm.30 An effective response necessitates disclosure of information regarding the breach to those individuals affected by it, as well as to persons and entities in a position to cooperate, either by assisting in notification to affected individuals or playing a role in preventing or minimizing harms from the breach.

b. Disclosure of Information. Often, the information to be disclosed to such persons and entities is maintained by federal agencies and is subject to the Privacy Act (5 U.S.C. § 552a). The Privacy Act prohibits the disclosure of any record in a system of records by any means of communication to any person or agency absent the written consent of the subject individual, unless the disclosure falls within one of twelve statutory exceptions.31 In order to ensure an agency is in the best position to respond in a timely and effective manner, in accordance with 5 U.S.C. § 552a(b)(3) of the Privacy Act, agencies should publish a routine use for appropriate systems specifically applying to the disclosure of information in connection with response and remedial efforts in the event of a data breach as follows:

To appropriate agencies, entities, and persons when (1) [the agency] suspects or has confirmed that the security or confidentiality of information in the system of records has been compromised; (2) the Department has determined that as a result of the suspected or confirmed compromise there is a risk of harm to economic or property interests, identity theft or fraud, or harm to the security or integrity of this system or other systems or programs (whether maintained by the Department or another agency or entity) that rely upon the compromised information; and (3) the disclosure made to such agencies, entities, and persons is reasonably necessary to assist in connection with the Department’s efforts to respond to the suspected or confirmed compromise and prevent, minimize, or remedy such harm.32

As described in the President’s Identity Theft Task Force’s Strategic Plan, all agencies should publish a routine use for their systems of records allowing for the disclosure of information in the course of responding to a breach of federal data.33 Such a routine use will serve to protect the interests of the individuals whose information is at issue by allowing agencies to take appropriate steps to facilitate a timely and effective response, thereby improving their ability to prevent, minimize, or remedy any harm resulting from a compromise of data maintained in their systems of records.

Attachment 3: External Breach Notification

To ensure consistency across government, this Attachment identifies the questions and factors each agency should consider in determining when notification outside the agency should be given and the nature of the notification.34 This Attachment does not attempt to set a specific threshold for external notification since breaches are specific and context dependant and notification is not always necessary or desired. The costs of any notifications must be borne by the agency experiencing the breach from within existing resources.

A. Background

1. Harm. Breaches can implicate a broad range of harms to individuals, including the potential for identity theft; however, this Section does not discuss actions to address possible identity theft or fraud. Agencies are referred to the ID Theft Task Force’s Strategic Plan for guidance.

2. Requirement. Agencies must implement the one specific new requirement discussed below; i.e., develop a breach notification policy and plan (see Section B. below).

3. Threshold questions. Both the decision to provide external notification on the occasion of a breach and the nature of the notification will require agencies to resolve a number of threshold questions.35 The likely risk of harm and the level of impact will determine when, what, how and to whom notification should be given.36

Notification of those affected and/or the public allows those individuals the opportunity to take steps to help protect themselves from the consequences of the breach. Such notification is also consistent with the “openness principle” of the Privacy Act that calls for agencies to inform individuals about how their information is being accessed and used, and may help individuals mitigate the potential harms resulting from a breach.

4. Chilling Effects of Notices.

A number of experts have raised concerns about unnecessary notification and the chilling effect this may have on the public.

In addition, agencies should consider the costs to individuals and businesses of responding to notices where the risk of harm may be low. Agencies should exercise care to evaluate the benefit of notifying the public of low impact incidents.

B. New Requirement

Each agency should develop a breach notification policy and plan comprising the elements discussed in this Attachment. In implementing the policy and plan, the Agency Head will make final decisions regarding breach notification.

Six elements should be addressed in the policy and plan and when considering external notification:

• whether breach notification is required

• timeliness of the notification

• source of the notification

• contents of the notification

• means of providing the notification

• who receives notification: public outreach in response to a breach

To ensure adequate coverage and implementation of the plan, each agency should establish an agency response team including the Program Manager of the program experiencing the breach, Chief Information Officer, Chief Privacy Officer or Senior Official for Privacy, Communications Office, Legislative Affairs Office, General Counsel and the Management Office which includes Budget and Procurement functions.38 A more detailed description of these elements is set forth below:

1. Whether Breach Notification is Required

To determine whether notification of a breach is required, the agency should first assess the likely risk of harm caused by the breach and then assess the level of risk. Agencies should consider a wide range of harms, such as harm to reputation and the potential for harassment or prejudice, particularly when health or financial benefits information is involved in the breach.

Agencies should bear in mind that notification when there is little or no risk of harm might create unnecessary concern and confusion.

Additionally, under circumstances where notification could increase a risk of harm, the prudent course of action may be to delay notification while appropriate safeguards are put in place.

Five factors should be considered to assess the likely risk of harm:

a. Nature of the Data Elements Breached. The nature of the data elements compromised is a key factor to consider in determining when and how notification should be provided to affected individuals.41 It is difficult to characterize data elements as creating a low, moderate, or high risk simply based on the type of data because the sensitivity of the data element is contextual. A name in one context may be less sensitive than in another context.42 In assessing the levels of risk and harm, consider the data element(s) in light of their context and the broad range of potential harms flowing from their disclosure to unauthorized individuals.

b. Number of Individuals Affected. The magnitude of the number of affected individuals may dictate the method(s) you choose for providing notification, but should not be the determining factor for whether an agency should provide notification.

c. Likelihood the Information is Accessible and Usable. Upon learning of a breach, agencies should assess the likelihood personally identifiable information will be or has been used by unauthorized individuals. An increased risk that the information will be used by unauthorized individuals should influence the agency’s decision to provide notification.

The fact the information has been lost or stolen does not necessarily mean it has been or can be accessed by unauthorized individuals, however, depending upon a number of physical, technological, and procedural safeguards employed by the agency. (See Attachment 1 above.) If the information is properly protected by encryption, for example, the risk of compromise may be low to non-existent.43

Agencies will first need to assess whether the personally identifiable information is at a low, moderate, or high risk of being compromised. The assessment should be guided by NIST

security standards and guidance. Other considerations may include the likelihood any unauthorized individual will know the value of the information and either use the information or sell it to others.

d. Likelihood the Breach May Lead to Harm

1. Broad Reach of Potential Harm. The Privacy Act requires agencies to protect against any anticipated threats or hazards to the security or integrity of records which could result in “substantial harm, embarrassment, inconvenience, or unfairness to any individual on whom information is maintained.”44 Additionally, agencies should consider a number of possible harms associated with the loss or compromise of information. Such harms may include the effect of a breach of confidentiality or fiduciary responsibility, the potential for blackmail, the disclosure of private facts, mental pain and emotional distress, the disclosure of address information for victims of abuse, the potential for secondary uses of the information which could result in fear or uncertainty, or the unwarranted exposure leading to humiliation or loss of self-esteem.

2. Likelihood Harm Will Occur. The likelihood a breach may result in harm will depend on the manner of the actual or suspected breach and the type(s) of data involved in the incident. Social Security numbers and account information are useful to committing identity theft, as are date of birth, passwords, and mother’s maiden name. If the information involved, however, is a name and address or other personally identifying information, the loss may also pose a significant risk of harm if, for example, it appears on a list of recipients patients at a clinic for treatment of a contagious disease.

In considering whether the loss of information could result in identity theft or fraud, agencies should consult guidance from the Identity Theft Task Force.45

e. Ability of the Agency to Mitigate the Risk of Harm. Within an information system, the risk of harm will depend on how the agency is able to mitigate further compromise of the system(s) affected by a breach. In addition to containing the breach, appropriate countermeasures, such as monitoring system(s) for misuse of the personal information and patterns of suspicious behavior, should be taken.46 Such mitigation may not prevent the use of the personal information for identity theft, but it can limit the associated harm. Some harm may be more difficult to mitigate than others, particularly where the potential injury is more individualized and may be difficult to determine.

2. Timeliness of the Notification

Agencies should provide notification without unreasonable delay following the discovery of a breach, consistent with the needs of law enforcement and national security and any measures necessary for your agency to determine the scope of the breach and, if applicable, to restore the reasonable integrity of the computerized data system compromised.

Decisions to delay notification should be made by the Agency Head or a senior-level individual he/she may designate in writing. In some circumstances, law enforcement or national security considerations may require a delay if it would seriously impede the investigation of the breach or the affected individual. However, any delay should not exacerbate risk or harm to any affected individual(s).

3. Source of the Notification

In general, notification to individuals affected by the breach should be issued by the Agency Head, or senior-level individual he/she may designate in writing, or, in those instances where the breach involves a publicly known component of an agency, such as the Food and Drug Administration or the Transportation Security Administration, the Component Head. This demonstrates it has the attention of the chief executive of the organization. Notification involving only a limited number of individuals (e.g., under 50) may also be issued jointly under the auspices of the Chief Information Officer and the Chief Privacy Officer or Senior Agency Official for Privacy. This approach signals the agency recognizes both the security and privacy concerns raised by the breach.

When the breach involves a Federal contractor or a public-private partnership operating a system of records on behalf of the agency, the agency is responsible for ensuring any notification and corrective actions are taken. The roles, responsibilities, and relationships with contractors or partners should be reflected in your breach notification policy and plan, your system certification and accreditation documentation, and contracts and other documents.

4. Contents of the Notification

The notification should be provided in writing and should be concise, conspicuous, plain language. The notice should include the following elements:

• A brief description of what happened, including the date(s) of the breach and of its discovery;

• To the extent possible, a description of the types of personal information involved in the breach (e.g., full name, Social Security number, date of birth, home address, account number, disability code, etc.);

• A statement whether the information was encrypted or protected by other means, when determined such information would be beneficial and would not compromise the security of the system;

• What steps individuals should take to protect themselves from potential harm, if any;

• What the agency is doing, if anything, to investigate the breach, to mitigate losses, and to protect against any further breaches; and

• Who affected individuals should contact at the agency for more information, including a toll-free telephone number, e-mail address, and postal address.

Given the amount of information required above, you may want to consider layering the information as suggested in Section 5 below, providing the most important information up front, with the additional details in a Frequently Asked Questions (FAQ) format or on your web site. If you have knowledge the affected individuals are not English speaking, notice should also be provided in the appropriate language(s). You may seek additional guidance on how to draft the notice from the Federal Trade Commission, a leader in providing clear and understandable notices to consumers, as well as from communication experts who may assist you in designing model notices.47 A standard notice should be part of your approved breach plan.

5. Means of Providing Notification

The best means for providing notification will depend on the number of individuals affected and what contact information is available about the affected individuals. Notice provided to individuals affected by a breach should be commensurate with the number of people affected and the urgency with which they need to receive notice. The following examples are types of notice which may be considered.

a. Telephone. Telephone notification may be appropriate in those cases where urgency may dictate immediate and personalized notification and/or when a limited number of individuals are affected. Telephone notification, however, should be contemporaneous with written notification by first-class mail.

b. First-Class Mail. First-class mail notification to the last known mailing address of the individual in your agency’s records should be the primary means notification is provided. Where you have reason to believe the address is no longer current, you should take reasonable steps to update the address by consulting with other agencies such as the US Postal Service. The notice should be sent separately from any other mailing so that it is conspicuous to the recipient. If the agency which experienced the breach uses another agency to facilitate mailing (for example, if the agency which suffered the loss consults the Internal Revenue Service for current mailing addresses of affected individuals), care should be taken to ensure the agency which suffered the loss is identified as the sender, and not the facilitating agency. The front of the envelope should be labeled to alert the recipient to the importance of its contents, e.g., “Data Breach Information Enclosed” and should be marked with the name of your agency as the sender to reduce the likelihood the recipient thinks it is advertising mail.

c. E-Mail. E-mail notification is problematic, because individuals change their e-mail addresses and often do not notify third parties of the change. Notification by postal mail is preferable. However, where an individual has provided an e-mail address to you and has expressly given consent to e-mail as the primary means of communication with your agency, and no known mailing address is available, notification by e-mail may be appropriate. E-mail notification may also be employed in conjunction with postal mail if the circumstances of the breach warrant this approach. E-mail notification may include links to the agency and http://www.USA.gov48 web sites, where the notice may be “layered” so the most important summary facts are up front with additional information provided under link headings.

d. Existing Government Wide Services. Agencies should use Government wide services already in place to provide support services needed, such as USA Services, including toll free number of 1-800-FedInfo and

e. Newspapers or other Public Media Outlets. Additionally, you may supplement individual notification with placing notifications in newspapers or other public media outlets. You should also set up toll-free call centers staffed by trained personnel to handle inquiries from the affected individuals and the public.

f. Substitute Notice. Substitute notice in those instances where your agency does not have sufficient contact information to provide notification. Substitute notice should consist of a conspicuous posting of the notice on the home page of your agency’s web site and notification to major print and broadcast media, including major media in areas where the affected individuals reside. The notice to media should include a toll-free phone number where an individual can learn whether or not his or her personal information is included in the breach.

g. Accommodations. Special consideration to providing notice to individuals who are visually or hearing impaired consistent with Section 508 of the Rehabilitation Act of 1973 should be given. Accommodations may include establishing a Telecommunications Device for the Deaf (TDD) or posting a large type notice on the agency web site.

6. Who Receives Notification: Public Outreach in Response to a Breach

a. Notification of Individuals. The final consideration in the notification process when providing notice is to whom you should provide notification: the affected individuals, the public media, and/or other third parties affected by the breach or the notification. Unless notification to individuals is delayed or barred for law enforcement or national security reasons, once it has been determined to provide notice regarding the breach, affected individuals should receive prompt notification.

b. Notification of Third Parties including the Media. If communicating with third parties regarding a breach, agencies should consider the following.

1. Careful Planning. An agency’s decision to notify the public media will require careful planning and execution so that it does not unnecessarily alarm the public. When appropriate, public media should be notified as soon as possible after the discovery of a breach and the response plan, including the notification, has been developed. Notification should focus on providing information, including links to resources, to aid the public in its response to the breach. Notification may be delayed upon the request of law enforcement or national security agencies as described above in Section 2. To the extent possible, when necessary prompt public media disclosure is generally preferable because delayed notification may erode public trust.

2. Web Posting. Agencies should post information about the breach and notification in a clearly identifiable location on the home page of your agency web site as soon as possible after the discovery of a breach and the decision to provide notification to the affected individuals. The posting should include a link to Frequently Asked Questions (FAQ) and other talking points to assist the public’s understanding of the breach and the notification process.49 The information should also appear on the web site. You may also consult with GSA’s USA Services regarding using their call center.

3. Notification of other Public and Private Sector Agencies.

Other public and private sector agencies may need to be notified on a need to know basis, particularly those that may be affected by the breach or may play a role in mitigating the potential harms stemming from the breach.

4. Congressional Inquiries. Agencies should be prepared to respond to inquires from other governmental agencies such as the Government Accountability Office and Congress.

c. Reassess the Level of Impact Assigned to the Information. After evaluating each of these factors, you should review and reassess the level of impact you have already assigned to the information using the impact levels defined by the NIST.51 The impact levels – low, moderate, and high, describe the (worst case) potential impact on an organization or individual if a breach of security occurs.52

• Low: the loss of confidentiality, integrity, or availability is expected to have a limited adverse effect on organizational operations, organizational assets or individuals

• Moderate: the loss of confidentiality, integrity, or availability is expected to have a serious adverse effect on organizational operations, organizational assets or individuals.

• High: the loss of confidentiality, integrity, or availability is expected to have a severe or catastrophic adverse effect on organizational operations, organizational assets or individuals.

The impact levels will help determine when and how notification should be provided. Where there is a range of risk levels attributed to the factors, the decision to provide notification should give greater weight to the likelihood the information is accessible and usable and whether the breach may lead to harm. If agencies appropriately apply the five risk factors discussed in section 1 of this attachment within the fact-specific context, it is likely notification will only be given in those instances where there is a reasonable risk of harm and will not lead to the overuse of notification.

Attachment 4: Rules and Consequences

A. New Requirement: Rules and Consequences Policy.

Fairness requires that managers, supervisors and employees be informed and trained regarding their respective responsibilities relative to safeguarding personally identifiable information and the consequences and accountability for violation of these responsibilities. Therefore, it is the responsibility of each agency head to develop and implement an appropriate policy outlining the rules of behavior and identifying consequences and corrective actions available for failure to follow these rules. Consequences should be commensurate with level of responsibility and type of personally identifiable information involved. Supervisors also must be reminded of their responsibility to instruct, train and supervise employees on safeguarding personally identifiable information. Agencies should develop and implement these policies in accordance with the agency’s respective existing authorities.

As with any disciplinary action, the particular facts and circumstances, including whether the breach was intentional, will be considered in taking appropriate action. Supervisors also should be reminded that any action taken must be consistent with law, regulation, applicable case law, and any relevant collective bargaining agreement. Supervisors should understand they may be subject to disciplinary action for failure to take appropriate action upon discovering the breach or failure to take required steps to prevent a breach from occurring.

Agencies having questions regarding development of a rules and consequences policy may contact OPM’s Center for Workforce Relations and Accountability Policy at (202) 606-2930.

1. Affected Individuals. At a minimum, each agency should have a documented policy in place which applies to employees of the agency (including managers), and its contractors, licensees, certificate holders, and grantees.

2. Affected Actions. The agency’s policy should describe the terms and conditions affected individuals shall be subject to and identify available corrective actions. Rules of behavior and corrective actions should address the following:

• Failure to implement and maintain security controls, for which an employee is responsible and aware, for personally identifiable information regardless of whether such action results in the loss of control53 or unauthorized disclosure of personally identifiable information;

• Exceeding authorized access to, or disclosure to unauthorized persons of, personally identifiable information;

• Failure to report any known or suspected loss of control or unauthorized disclosure of personally identifiable information; and

• For managers, failure to adequately instruct, train, or supervise employees in their responsibilities.

3. Consequences. Applicable consequences may include reprimand, suspension, removal, or other actions in accordance with applicable law and agency policy. The minimum consequence agencies should consider is prompt removal of authority to access information or systems from individuals who demonstrates egregious disregard or a pattern of error in safeguarding personally identifiable information.


1 The term “personally identifiable information” refers to information which can be used to distinguish or trace an individual’s identity, such as their name, social security number, biometric records, etc. alone, or when combined with other personal or identifying information which is linked or linkable to a specific individual, such as date and place of birth, mother’s maiden name, etc.

2 Title III of the E-Government Act of 2002, Pub. L. No. 107-347.

3 5 U.S.C. § 552a.

4 Executive Order 13402 charged the Identity Theft Task Force with developing a comprehensive strategic plan for steps the federal government can take to combat identity theft, and recommending actions which can be taken by the public and private sectors. On April 23, 2007 the Task Force submitted its report to the President, titled “Combating Identity Theft: A Strategic Plan.” This report is available at

5 For the purposes of this policy, the term “breach” is used to include the loss of control, compromise, unauthorized disclosure, unauthorized acquisition, unauthorized access, or any similar term referring to situations where persons other than authorized users and for an other than authorized purpose have access or potential access to personally identifiable information, whether physical or electronic.

6 Agencies should use a best judgment standard to develop and implement a breach notification policy. Using a best judgment standard, the sensitivity of certain terms, such as personally identifiable information, can be determined in context. For example, an office rolodex contains personally identifiable information (name, phone number, etc.). In this context the information probably would not be considered sensitive; however, the same information in a database of patients at a clinic which treats contagious disease probably would be considered sensitive information. Similarly, using a best judgment standard, discarding a document with the author’s name on the front (and no other personally identifiable information) into an office trashcan likely would not warrant notification to US-CERT.

7 Terms not specifically defined within this Memorandum (e.g., sensitive) should be considered to reflect the definition found in a commonly accepted dictionary.

8 FISMA security requirements apply to Federal information and information systems, including both paper and electronic format.

9 A plan to review the controls for information systems not previously included in other security reviews must be addressed in the agency’s breach notification policy (e.g., timeframe for completion of review, etc.); however, completion of the review for those systems is not required to be finished within the 120-day timeframe for development of the policy.

10 In this policy, “access” means the ability or opportunity to gain knowledge of personally identifiable information.

11 For example, FISMA or associated standards, policies, or guidance issued by OMB or the National Institute of Standards and Technology (NIST).

12 This Memorandum, or its attachments, should not be read to mean an agency’s failure to implement one or more of the many provisions of FISMA or associated standards, policies, or guidance issued by OMB or the National Institute of Standards and Technology (NIST) would constitute less than adequate protections required by the Privacy Act of 1974.

13 5 U.S.C. § 552a.

14 5 U.S.C. § 552a (e)(10).

15 The Privacy Act requires agencies to “maintain all records which are used by the agency in making any determination about any individual with such accuracy, relevance, timeliness, and completeness as is reasonably necessary to assure fairness to the individual in the determination” in their systems of records. 5 U.S.C. § 552a(e)(5).

14 The Privacy Act requires agencies to publish a notice of any new or intended use of information maintained in a system of records in the Federal Register to provide an opportunity for the public to submit comments. 5 U.S.C. § 552a(e)(4). Agencies are also required to publish notice of any subsequent substantive revisions to the use of information maintained in the system of records. 5 U.S.C. § 552a(e)(11). OMB Circular A-130 (“Management of Federal Information Resources”) offers additional guidance on this issue. OMB Circular A-130, App. I, sec. 4.c.

17 44 U.S.C. 3544(b).

18 Agencies may schedule training to coincide with existing activities, such as ethics training. Communications and training related to privacy and security must be job-specific and commensurate with the employee’s responsibilities. The Department of Defense, the Office of Personnel Management, and the Department of State offer agencies a minimum baseline of security awareness training as part of the Information Systems Security Line of Business.

19 Agencies should also consider augmenting their training by using creative methods to promote daily awareness of employees’ privacy and security responsibilities, such as weekly tips, mouse pads imprinted with key security reminders, privacy screens for public use of laptops, and incentives for reporting security risks.

20 To the extent agencies are substantively performing these reviews, agencies should leverage these efforts to meet the new privacy requirements. This provision does not apply to apply to the accessioned holdings (archival records) held by the National Archives and Records Administration (NARA).

21 The Department of Defense and Intelligence Community establish their own policy and guidance for the security of their information systems. 44 U.S.C. 3543(c).

22 Agencies with questions addressing this assignment regarding the Paperwork Reduction Act of 1995 (44 U.S.C. 3501 et seq.) should contact their respective desk officer at the Office of Management and Budget.

23 See OMB Memo 06-16 “Protection of Sensitive Agency Information” (

24 See NIST’s website at for a discussion of the certified encryption products.

25 Non cabinet agencies should consult the equivalent of a Deputy Secretary.

26 44 U.S.C. § 3544(b)(7).

27 For additional information on NIST guidance and standards, see

28 See “Computer Security Incident Handling Guide: Recommendations of the National Institute of Standards and Technology” (

29 The responsibilities of US-CERT are outlined in 44 U.S.C. § 3546. Its complete set of operating procedures may be found on the US-CERT website ( Separate procedures are in place for the Department of Defense as identified in Directive O-8530-1 and all components report incidents to the Joint Task Force Global Network Operations (JTF-GNO), which, in turn, coordinates directly with the US-CERT.

30 Here, “harm” means damage, fiscal damage, or loss or misuse of information which adversely affects one or more individuals or undermines the integrity of a system or program.

31 5 U.S.C. §§ 552a(b)(1)-(12).

32 See Appendix B of the Identity Theft Task Force report (

33 Id.

34 These factors do not apply to an agency’s notification to US-CERT. Agencies must report all incidents – potential and confirmed – involving personally identifiable information to US-CERT.

35 Notice may not be necessary if, for example, the information is properly encrypted because the information would be unusable.

36 See OMB’s September 20, 2006 memorandum titled “Recommendations for Identity Theft Related Data Breach Notification” for information and recommendations for planning and responding to data breaches which could result in identity theft (

37 Federal Trade Commission, Prepared Statement of the Federal Trade Commission Before the Committee on Commerce, Science, and Transportation, U.S. Senate, on Data Breaches and Identity Theft (Washington, D.C.: June 16, 2005), p. 10. In this testimony, the Federal Trade Commission raised concerns about the threshold for which consumers should be notified of a breach, cautioning that too strict a standard could have several negative effects.

38 Non-Cabinet-level agencies should include their functional equivalent.

39 For reference, the express language of the Privacy Act requires agencies to consider a wide range of harms: agencies shall “establish appropriate administrative, technical and physical safeguards to insure the security and confidentiality of records and to protect against any anticipated threats or hazards to their security or integrity which could result in substantial harm, embarrassment, inconvenience, or unfairness to any individual on whom information is maintained.” 5 U.S.C. § 552a (e)(10).

40 Another consideration is a surfeit of notices, resulting from notification criteria which are too strict, could render all such notices less effective, because consumers could become numb to them and fail to act when risks are truly significant.

41 For example, theft of a database containing individuals’ names in conjunction with Social Security numbers, and/or dates of birth may pose a high level of risk of harm, while a theft of a database containing only the names of individuals may pose a lower risk, depending on its context.

42 For example, breach of a database of names of individuals receiving treatment for contagious disease may pose a higher risk of harm, whereas a database of names of subscribers to agency media alerts may pose a lower risk of harm.

43 In this context, proper protection means encryption has been validated by NIST.

44 5 U.S.C. § 552a(e)(10).

45 See “Recommendations for Identity Theft Related Data Breach Notification” (

46 For example, if the information relates to disability beneficiaries, monitoring a beneficiary database for requests for change of address may signal fraudulent activity.

47 Additional guidance on how to draft a notice is available in the FTC publication titled “Dealing with a Data Breach” ( Although the brochure is designed for private sector entities that have experienced a breach, it contains sample notice letters that could also serve as a model for federal agencies. You may also seek guidance from communications experts who may assist you in designing model notices.

48 The current domain name for the Federal Internet portal required by section 204 of the E-Government Act of 2002 is

49 See the FAQ posted by the Department of Veterans Affairs in response to the May 2006 incident for examples of links to identity theft resources and a sample FAQ (

50 For example, a breach involving medical information may warrant notification of the breach to health care providers and insurers through the public or specialized health media, and a breach of financial information may warrant notification to financial institutions through the federal banking agencies.

51 See FIPS 199 and Attachment 1 of this memorandum. Reassessment is suggested as the context of any breach may alter your original designation.

52 The determination of the potential impact of loss of information is made by the agency during an information system’s certification and accreditation process.

53 Here, “control” means the authority of the government agency that originates information, or its successor in function, to regulate access to the information. Having control is a condition or state and not an event. Loss of control is also a condition or state which may or may not lead to an event, i.e., a breach.

FISMA Privacy Management and Implementation

Annual Reporting Instructions for the Federal Information Security Management Act (FISMA) and Agency Privacy Management – Annual memorandum released by the Office of Management and Budget to the heads of departments and agencies within the Federal Government providing instructions for agency reporting under the Federal Information Security Management Act of 2002. Declares that OMB will not ask for privacy related information in annual E-Government Act submissions.

OMB Circular A-130, Management of Federal Information Resources (November 28, 2000) (23 pp. PDF) – Memorandum released by the Office of Management and Budget to the heads of executive departments and agencies within the Federal Government providing policy for the management of Federal information resources. Includes procedural and analytic guidelines for implementing specific aspects of these policies.

OMB M-99-05, Instructions on Complying With President’s Memorandum of May 14, 1998, “Privacy and Personal Information in Federal Records” (January 7, 1999) – Memorandum released by the Office of Management and Budget to the heads of departments and agencies within the Federal Government providing instructions on complying with the President’s Memorandum of May 14, 1998, on “Privacy and Personal Information in Federal Records.”

Freedom of Information Act (FOIA)

Presidential Memoranda issued January 21, 2009 – provides direction to Federal agencies on the FOIA. “All agencies should adopt a presumption in favor of disclosure, in order to renew their commitment to the principles embodied in FOIA, and to usher in a new era of open Government.”

E-Government Act of 2002

OMB M03-02, OMB Guidance for Implementing the Privacy Provisions of the E-Government Act of 2002 (September 30, 2003) – Memorandum released by the Office of Management and Budget to the heads of executive departments and agencies within the Federal Government providing instructions for implementing the Privacy Provisions of the E-Government Act of 2002. Instructs on privacy protections “when Americans interact with their government.”

Privacy Act Guidance

The Attorney Generals’ Guidelines for Domestic FBI Operations (September 2008) (46 pp. PDF) – applies to the “investigative activities conducted by the [Federal Bureau of Investigation] FBI within the United States or outside the territories of all countries.” “The FBI may provide investigative assistance to state, local, or tribal agencies in the investigation of matters that may involve federal crimes or threats to the national security….” The Guidance addresses the Privacy Act specifically.

  • FBI Fact sheet on the “new” (2008) consolidated Guidelines
  • “The Privacy Act restricts the maintenance of records relating to certain activities of individuals who are United States persons, with exceptions for circumstances in which the collection of such information is pertinent to and within the scope of an authorized law enforcement activity or is otherwise authorized by statute. 5 U.S.C. 552a(e)(7). Activities authorized by these Guidelines are authorized law enforcement activities or activities for which there is otherwise statutory authority for purposes of the Privacy Act. These Guidelines, however, do not provide an exhaustive enumeration of authorized FBI law enforcement activities….”

DOJ Overview of the Privacy Act of 1974
(May 2004) – Department of Justice (DOJ) discussion includes citations to court decisions interpreting agency Privacy Act of 1974 data quality requirements.

OMB Privacy Act Guidance—Update (May 24, 1985) (10 pp. PDF) – Document from the Office of Management and Budget for senior agency officials identifying “three areas in which agencies should amend their practices to new interpretations of the Privacy Act which have resulted from recent Congressional action or judicial interpretation.” Suggestions affect Section 6 of the Privacy Act of 1974 and were created to help agencies implement the Act’s provisions.

Implementation of the Privacy Act of 1974, Supplementary Guidance, 40 Fed. Reg. 5674, (December 4, 1975) (3 pp. PDF) – Guidance from the Office of Management and Budget to the Heads of Federal Executive Departments and Establishments regarding “comments and questions of general interest” raised in the wake of the release of the Privacy Act of 1974.

OMB Privacy Act Implementation, Guidelines and Responsibilities, 40 Fed. Reg. 28948, 28965 (July 9, 1975) (32 pp. PDF) – Memorandum from the Office of Management and Budget to the Heads of Executive Departments and Establishments regarding legislation implemented to insure that Federal agencies protect individual privacy rights when collecting personal information.

Computer Matching

Final Guidance Interpreting the Provisions of Public Law 100-503, the Computer Matching and Privacy Protection Act of 1988, 54 Fed. Reg. 25818 (June 16, 1989) (12 pp. PDF) – Guidance issued by the Office of Management and Budget regarding interpretation of the provisions of Public Law 100-503, and the Computer Matching and Privacy Protection Act of 1988. Discusses procedural safeguards affecting agencies’ use of Privacy Act records in performing certain types of computerized matching programs.